Visit to discover Indian blogs

Tuesday, August 22, 2017

Poor Economics: lessons for government, NGO's and us

A lot of us want to do something to eradicate poverty from the world. However, monetary charity may not always help since people often lack the discipline and knowledge required to fulfil a long term vision of getting out of poverty and in some cases the money might be taken away by someone more powerful (such as a political elite) who may use it in his own interest. This is consistent with behavioural research which has highlighted that people don’t necessarily make decisions in their own interest. Also Abhijit Banerjee and Esther Duflo note that poor are not able to get out of poverty partially because they are incharge of too many decisions for themselves. These can involve decisions as simple as mixing chlorine in water or choosing immunisation. Individuals are also risk averse. When we correct for the risk aversion at the individual level, society will benefit on the whole with certain impositions even though the outcomes are not known with certainty at the individual level. This is why many decisions that we have to make for ourselves are imposed upon us without our knowledge.

In Poor Economics, Banerjee and Duflo try to arrive at the most efficient interventions in terms of benefits relative to cost by focussing on research that uses randomised controlled trials (RCTs) and other quantitative research methods. I summarise these in broad categories of health, education, fertility, micro finance and institutions. All of these are interrelated and improvements in one area often spread to other dimensions as well.

The old adage prevention is better than cure fits well in taking steps to counter poverty. It has been established through the RCTs that relatively smaller interventions for the poor such as Iron capsules to prevent anaemia, distribution of insecticide treated mosquito nets for malaria and deworming in Kenya have a significant impact on future monetary income that far outweighs the costs. These interventions also have an intrinsic value because being healthy directly improves the quality of life. Other minor interventions that can benefit health of the poor greatly are the use of chlorinated water or availability of ORS for treating diarrhoea. In a similar vein, using soaps can prevent a large number of individuals from hygiene related infections. One major issue that all the developing countries have faced is the low rate of vaccination among the poor. People are usually shortsighted and only see immediate benefit of preventive health services. Banerjee and Duflo argue that incentivizing people through small gifts or cash transfers reduces the average cost of immunization per person given the fixed cost involved with health camps. The benefits accrued from additional cost may also be obtained by government in the form reduced health expenditure on curing these individuals in public hospitals. Such a policy must be seen as investment in health which is a genuine form of human capital, resulting in faster growth and possibly increased tax receipts by the virtue of higher future income of the beneficiaries. An individual who is healthy will also lead to improved health for others around him. Benefits of being healthy extend beyond the people an individual cares about. This strengthens the case for public intervention in health. 

Most of us find it psychologically difficult to move away from our habits. And it is perhaps too much to ask a person of limited means to spend his money to experiment on something new. TAMTAM experiments in Africa have demonstrated that people who were given free insecticide treated mosquito bednets once were more likely to purchase them at the market price in future. This another good argument for giving a one time subsidy on certain goods to poor to nudge them toward better habits.

Considering both demand as well as supply side factors are important for improving educational outcomes. Telling parents returns on education is the most effective way to make them send their kids to school. Parents assume that at lower levels of education, an additional year of schooling has no monetary returns and there are disproportionately large returns at higher levels of education. However, this is not the case. An additional year of schooling adds to income in the same proportion at each level of schooling. It needs to be noted however, that if the returns to education are proportional then there indeed is an education based poverty trap as each additional year of schooling adds a larger amount to the income in absolute terms. This means it might make sense for parents to put the eggs in one basket and educate only single child that they perceive intelligent. 

We can also provide group loans of a long term nature to the poor to educate their children at primary level. This might also have the benefit that people will monitor each other's children and also see that schools are doing their jobs. It has also been found from Mexico’s Progresa program that cash transfers to women improved child education outcomes. This is obvious since women are more altruistic toward their children and making transfers to them improves their decision making power within the household. The market for education is just like any other market where the consumers have no idea about the quality of the product that they are getting. The consequences are quite easily foreseeable. The children cannot verify if the poor results are because of themselves or because of teaching and tend to lose confidence in their ability to understand. The lack of confidence leads to loss of interest which leads to low learning outcomes. And so the following improvements are required in the method of teaching:
1)  The children need to be encouraged constantly.
2)  The pace should be according to their pace of understanding for each individual student
3)  Stigma with repeating grades should be removed
4)  Teachers have to be as enthusiastic about weak students and not create an elitist environment (which is the usual practice)
5)  Computer based learning to be promoted
6)  Division of classes according to performance essential and also continuous evaluation is needed
7)  Community based learning is required
8)  We also need to avoid priming characteristics such as caste, religion etc.
9)  Deferred payment mechanisms by NGO's might be the way forward
10) Government policy should incentivise education through conditional cash transfers

Female fertility
Empirically, the number of children has been found to have no adverse impact on children's education. However, it has a negative impact on mother's well being. Since, childbearing is costly for women, it makes economic sense for women to prefer less children compared to men on an average. Therefore, to improve life of women, the best policy for reducing fertility is to improve intra-household female bargaining power (e.g. through property titles), affecting social norms (e.g. this has happened through soap opera in brazil), contraceptive distribution in the absence of male partner and increasing their level of schooling. One of the motives for having children is old age security. Thus, social security and financial development can also reduce the fertility rate by reducing the need for children.

Microfinance and savings
Poor face a number of risk (of health, crop failure etc.). They mitigate these risks using certain steps that include but are not limited to:
1) Doing multiple jobs
2) Those working in agricultural sector diversify cropping land
3) Helping each other out and marriages done in neighbouring villages as an insurance in case one of the villages suffers from a calamity (you can also think how this influences the norm of not marrying within village.).

Risk prevents poor farmers in villages exposed to more risk  (e.g. rainfall risk in non irrigated places) to invest in more efficient technology. People can and do reduce risk for each other by entering a contract (formal/informal) to help each other out in case of a calamity. Traditions sometimes however prevent from giving the most efficient help instead of unneeded things (e.g. in marriages, sickness etc.). Insurance is one such formal contract. However, insurance is marred with moral hazard as well as adverse selection problem which can be solved by providing insurance against major health problems and agricultural insurance based on rainfall. Coupling microfinance with insurance can potentially reduce loan default risk for microfinance institutions too. Lending to poor has higher default risk, difficult to monitor, collection difficult (e.g. Intimidating looking Kabulwalas were often hired in West Bengal in 70s and 80s as collection agents). It is also tough to leave a given moneylender so lenders exercise monopoly power over borrowers. Banks often do not give loans to poor. A partial solution to this problem is group lending through microfinance which solves moral hazard as well as adverse selection. As Banerjee and Duflo note, micro finance institutions have often received undue criticism. A case in point is media reports criticising SKS micro finance for suicide of  57 farmers Andhra Pradesh in 2010. However, these farmers did not actually take loans from SKS. This is not to say that micro finance is a magical tool that will solve the problem of poverty. There are certain fundamental problems with microcredit:
1) Risk aversion
2) Non inclusion of new members in group
3) Lack of flexibility (so that longer term and larger loans cannot be disbursed)

Saving more in each period can bring the poor out of poverty trap. They find it difficult to save more because of pull of instant gratification. Somewhat bizarrely, Banerjee and Duflo report that some women borrowed from spandana at 24% rate and deposited the sum at 4% rate in savings account implying they are willing to pay some amount to make them exercise self control. At low levels of consumption, marginal utility of consumption is high and it is difficult not to spend money at hand. Just like it's difficult to refrain from sweets when they are in front ("temptation goods"). This also might explain healthier food choices made by rich. We don’t necessarily look at future consequences. This brings in the need for a nudge such as savings account which have been reported to make people save more money. 

Availability of fertilizers at doorstep immediately after harvest, when farmers are less cash constrained, may prompt farmers to buy these which will lead to increase in productivity. Another nudge may be to give fertilizer voucher for purchase at a later time (similar to SAFI programme in Kenya). An addition to the policy can be that when a person delays voucher redemption, more fertiliser will be provided on the voucher (upto a certain date). This is similar to interest on holding fertilizer voucher. It should also be ensured that productivity improving products are readily available. Additionally, individuals will find other alternatives when an emergency situation arises as returns to fertilizer use are very high (around 70%). Similar instruments can be designed for the temptation goods.

Aspirations and institutions
Aspirations play an important role in determining the life trajectory of an individual. The poor usually aspire for their children to get a secure government job as business involves risk, time, effort and is difficult to start since capital is lumpy. Therefore, permanent jobs with housing plans may help the poor in migrating to the cities and get a better education. Manufacturing growth has been more important than green revolution in raising rural wages. Labourers with a low skill level have also gained from this. Therefore, outward development strategy of rural areas may turn out to be more useful. To break the vicious circle where lack of jobs do not allow for improving worker quality and which prevents setting up of industries in rural areas, a strategy of investment in rural infrastructure is what is needed. The policy implications for this are skill training for better jobs and tax incentives for factories requiring such labour if they set up in rural areas. 

From a regulatory point we face two major challenges. First issue is the labour laws which are counterproductive and harm the workers more than benefitting them. These laws need to be simplified. The second issue is the discrimination in jobs which creates a double disadvantage for women as it also leads to discrimination in education for women and affects parental motivation in investing in girl child's education. But first of all we need better estimates of the magnitude of job discrimination against women (after taking into account differences in skill, schooling, effort etc.).

Institutions, loosely described as legal and cultural factors, are important for development as well. For example, the districts which had zamindari system during British rule in India are much poorer with lesser productivity today. However a series of small changes can have big effects. A study in Uganda showed that once the corruption in school funds was exposed, the proportion of funds actually reaching the school increased from 13% to 80% as it was riskier to get caught. Major changes, at least in India have been motivated from within the system. For example RTI has become a powerful tool to check corruption. Women’s representation in politics is also crucial. Gender based reservations in gram panchayats in india have reported to generate investment in different sorts of infrastructure (e.g. towards schooling). Integration of arts with development issues can have its own contribution. Awareness campaigns such as street plays can be a useful tool in making people vote on development issues rather than ethnic favorites. Improving institutions also requires disseminating information about agenda of political parties and its compliance once they come to power. 

Ultimately, attention to detail, rational thinking, willingness to experiment and optimism are all necessary ingredients to improving developmental outcomes and reducing poverty!